Re: [femm] resistive losses and Lubarsky

[David Meeker <dmeeker@xxxxxxxx>]

Henk Wensink wrote:
> Dear David Meeker,
>
>
> Before I start complaining, thanks for your program. It was exactly what
> I
> needed.
>
> One of my objectives was to calculate the resistance at high frequencies.
> Message number 49 of the "femm mailing list" got me a bit worried whether
> or
> not the eddy currents were already included in the resistive losses when
> using the last FEMM release. In the simulations I did, this component seems
> indeed been accounted for. However, in the manual it is stated that:
> "Resistive losses; This selection integrates the i^2 R losses due to
> currents flowing in the “z” direction (or theta direction, if you are
> evaluating an axisymmetric problem)."
>
> Which sounds to me that eddy currents (that are not flowing in the theta
> direction) are not included! Could you please clarify this point?
If the problem is really axisymmetric, all of the eddy currents flow in the
into-the-page (i.e. theta) direction.  The "resistive losses" integral does
include the resistive losses due to the eddy currents.
> Second, it seems I found a bug. When evaluating in the FEMM-editor, in the
> LUA console window "runpost("temp.fem")", it does run the FEMM-viewer, but
> it gives a "run error" (the FEMM-file is attached to this mail.).
> With another file, it first states:
> "
> error: <statement> expected;
> last token read: '[' at line 1 in file 'temp.fem'
> "
> after that, it gives the "run error".
> After clicking OK to these messages, the FEMMviewer is working normal.
>
> When using the appropriate button in the FEMM-editor to invoke the
> FEMM-viewer, there are no errors for both files.
When you do the "runpost" command, the argument that you give it isn't supposed
to be a .fem file.  The postprocessor is expecting to be pointed to a lua
script.  Instead of trying to run a lua script, it is trying to parse the
.fem file that you'd specified, and it throws an error.
> My third remark. Am I right to think that there is no way for the LUA script
> to "switch" from the Lua window in the editor to the Lua window in the
> viewer? So a single script can not invoke the design, analyse it and do
> calculations on it as well.
This is still something of a kludge.  You can't do what you are after with
a single script, but you can do it with two scripts.  You have on preprocessor
script that builds the geometry, runs the analysis, and calls the postprocessor
and tells it what script to run.  The postprocessor script then performs
whatever analysis is required, then possibly writing the results to disk.
 The preprocessor script could then read the results from the file on disk
to use the results in its next iteration, or whatever.
> And my final remark. The manual states that FEMM can be used for solving
> low
> frequencies problems. My simulations are run at several hundredths MHz, and
> the calculated values (inductance, resistance)seem pretty accurate. Can you
> give a general idea what is a high or a low frequency?
"Low frequency" means problems that can be described by a diffusion equation
rather than by a wave equation. There are some metrics that you can look
at--basically the wavelength at the frequency of interest must be much longer
than some characteristic dimensions of your problem.  You might also check
out http://web.mit.edu/6.013_book/www/chapter15/15.3.html 
If all the parts in the problem are very small, the frequencies at which
the "low frequency" formulation is valid could be pretty high.

Dave.